70 thoughts on “CCD… announcement coming soon!

  1. Something better than M9? Megapixel count is meaningless, dynamic range, tonality, color and noise are more important than just a resolution increase.

  2. This could well be the kind of news i’ve been waiting for ! Greater than 24MP, cool, but color (which is usually great with CCD) and high ISO perf (which is usually very bad with CCD) Matter a lot ! Anyway i’m very excités to know about it. Can’t wait. Thanks for the heads up-ish ! 😉

    1. Colours are a function of the colour filter array, not really the underlaying capturing device. If anything CMOS allows for better colour reproduction under limited SNR conditions.

  3. high ISO performance is actually quite good with a CCD, which is why they continue to be favored in scientific applications …their manufacture can be costly, but I’m hoping this comes through as I have been holding out

    1. They are often favored for other reasons – like availability, low (!) price for special products (they use old fabs already paid for), TDI (time delay integration), compability with existing systems and so on.

      The “high ISO” vs “low ISO” thing is not much more than a myth. If a sensor performs really well at low ISO it will also perform well at high ISO (though not necessarily as well as a sensor designed with high ISO in mind). This is because most noise by far is because of nature of light itself.

      (Though some cameras have poor ADCs which reduce the low ISO dynamic range, notably many Canons – nothin to do with CCD/CMOS though)

    1. Then again, unique shape may be code for “not Leica”. No Tomato Face? No up charge for Leica cache? If it’s a rangefinder my wallet is already open. 🙂

    1. Highest probability for Voigtlander, I guess. Zeiss Ikon rangefinder with its all time best finder ever would be great.

      1. Really ? Video Caméras with CCD sensors used to have ones, is that very different for a Stills camera (i’m really asking, to understand, not for just for the sake of internet debate/argument 😉

  4. That’s some exciting news! I’d love it to be a rangefinder and Voigtländer as Zeiss already have the knowledge. I shot the wonderful Zeiss Ikon for two years. But due to “unique shape” it might well be something “lcd only” with attachable optical finders. Probably to reduce cost. I don’t care very much for the megapixel count. Hopefully it’s at least coming with an EVF optimized for manual focus. Couldn’t they just take a bessa or Ikon and put the damn sensor in it..?

  5. I for one do not really hope for a Rangefinder, but an EVF. Or even better an hybrid à la Fuji X. Rangefinders are great, but easily broken or decalibrated. It they could come up with a new digital/hybrid rangefinding tech, that Would not rely so much on how well is your rangefinder mecanically adjusted but could tell when you actually are in focus, that would be stunning. EVFs have come a long way, and frankly are now very good. With a nice peaking / punch in zoom, you know you have it, no worries about focus. After all it will be digital, why not embrace it all the way.

  6. I just hope it has a rangefinder. Focusing M glass with an EVF is doable, but not nearly as enjoyable of a process…to me at least. I’m keeping my fingers crossed for a Zeiss/Voigtlander venture (although I know Zeiss has said multiple times a digital Ikon is not going to happen).

    1. You know what is even less enjoyable….the rangefinder going out of calibration. Seriously I love my M240 but EVF’s are only getting better and the benefits outway their downside IMO.

      1. There are several focus technologies out there that I am not able to explain in detail: it could happen with a second smaller cmos sensor that can be calibrated by the owner itself if necessary, there are laser or ultra sonic rangefinders and – as far a I know – Leica is also developing a digital-optical modern rangefinder for the next M. Therefore on sensor focusing is not mandatory for replacing the mechanical rangefinder, if the CCD sensor can not serve an EVF or if an optical finder is wanted. Mechanical rangefinders are surely outdated IMHO.

    1. This is false. Both CCD and CMOS just capture photons and convert them to electrons. The colour is defined by the colour filter array on top of the sensor. You can study the colour accuracy figures from DxOMark tests.

      1. It’s not true, i shoot since 87, CCD Sensors do have a better Color Rendition, period. Don’t come with these insane, unserious DxOMark Crap. Typical DSLR Sensors (that doesn’t mean 3 Color Foveon Sensors) simply do have the usual RGB-Bayern Demosaic, only Fujifilm does have their “X-Trans” Color Filter Array in Marketing Speech.

        Just google the now very old comparsion between the Pentax K-5 (CMOS) vs. *istDL (6MP, CCD) 1:1 Picture comparsion. For that, i do also love still my old D80, D60 & D200 compared to D90, D7000 & alikes.

        1. So scientific measurements are wrong and even “crap” because you feel diffrent?

          The RGB arrays are not identical to each other in their spectral response. Some are more strict in their colour separation, while others offer significat overlap – both have advantages.

          Pentax K-5 and the old *istDL have _different color filter arrays with different spectral responses_. Sure both are just RGBG Bayer pattern, but that doesn’t mean the spectral responses are the same.

          Here’s a practical test with the two most recent Pentax medium format digital cameras: http://www.dpreview.com/forums/post/57848263 – Practically identical for CMOS and CCD.

          Here is Eric Fossum, the inventor of APS (i.e. CMOS pixel) on this subject: http://www.dpreview.com/forums/post/54028441

          Why in your opinion would CCD be superior? What would be the technological reason?

    2. You are making a confusion between the sensor technology and the rendition chosen by manufacturers. Sensors, be they CCD or CMOS, only ‘see’ in black and white and that’s the RGB array, together with the software of the camera, that allows for colour rendition.

      1. That’s a gross oversimplification that leads to a false conclusion. You make the assumption that CCD and CMOS sensors have the same monochrome spectral response. But they don’t. Even if the naked single pixels of CCD and CMOS sensors are colourblind they still react differently to different spectral distributions, i.e. both types of sensors will produce different black and white representations of the same light they’re exposed to. Furthermore, the response of CCD sensors much better matches the response of the human eye than the one of CMOS sensors. And that’s not something that is easy to compensate for in software. The RGB array on top of the sensor is only one of many variables in the equation that results in the colour rendition of a sensor.

        1. That is just wrong CCD and CMOS both use the same mothod (photoelectric effect), and same material (silicon) to capture light. CCD does not have some mystical “better response”. This is not only a claim with zero hard evidence, but there is counter evidence, like this – Pentax 645 CCD vs Pentax 645 CMOS – pracitally identical colour response: http://www.dpreview.com/forums/post/57848263

          Eric Fossum, the inventor of APS (ie CMOS pixel) also has said this on the subject: http://www.dpreview.com/forums/post/54028441

          CCD colour advantage is just a myth.

  7. I doubt this camera will be made by Leica. Leica has previously stated that CMOS is the way forward and that a pixel is just a way of converting a photon into an electrical signal. And, the characteristics of the output is due to the processing.

    “Many people think there is a big difference in the touch and feel and the look and feel of the CCD vs CMOS. We think a pixel just renders light or transforms light into electricity. And the look and feel is done in the image processing. On the other hand, the CMOS sensors have a lot of advantages such as video and live view and we therefore think that the CMOS have the future at Leica.” – Stefen Daniel, February, 2014

  8. This is killing me, can’t wait. Somebody Please call me as soon as you have the preorder information! 867-5309


  9. There is no reason for any manufacturer to come up with a CCD based camera any more, just like there is no reason for car companies to come up with a steam engined vehicle. Simply CCD offers no tanglible advantages over CMOS nowdays, but CMOS offers far lower power consumption, superior live view, higher signal-to-noise ratio, larger dynamic range, better colour accuracy (due to larger SNR) in low light/shadows.

    When it comes to scientific CCDs – the main reason why they are used is the easy implementation of global shutter, though if it’s implemented, half of sensor area is sacrificed (lower sensitivity to light even with microlenses and lower full well capacity or the maximum signal, thus lower SNR, DR etc.). Global shutter is nowdays available for CMOS (though also with drawbacks, like larger noise) and slowly it’s eating this market away from CCD as well.

    1. Better color rendition. For some, that’s a tangible advantage that outweighs the only significant advantages of CMOS, high ISO performance and live view. Some people obviously can’t see this, that doesn’t mean it isn’t there.

      1. Colour difference is a myth.

        Both CCD and CMOS do the basic capturing process in identical way – turn photons into electrons. Electrons do not have colour. Which photons are allowed to create electrons is defined by the colour filter array on top of the sensor.

        Here’s a practical test with the two most recent Pentax medium format digital cameras: http://www.dpreview.com/forums/post/57848263 – Practically identical for CMOS and CCD. If there were difference, it would show here, yet it doesn’t

        Here is Eric Fossum, the inventor of APS (i.e. CMOS pixel) on this subject: http://www.dpreview.com/forums/post/54028441

        It’s not about “being able to see it”, but a matter of scientific fact that CCD and CMOS create the same colour performance (with possibly very slightly better response for the “blue photons” for the CMOS as per Fossum above).

        Now, in deeper shadow areas CMOS delivers clearer image which also means better output colour accuracy (as the SNR is higher).

        1. Hi Anu, To my eyes, m9 colours & blacks are different to my other CMOS based cameras (I think slightly cooler, blacker blacks & purples render better). It’s not much but definitely different. I wonder if part of this real world difference is the ways our eyes see these colours. I know older men especially, have a decaying colour spectrum, & maybe this ‘Difference’ I see is I either see ‘more’, or ‘less’ colour than other people, especially where CMOS vs CCD represent these ‘seen’ or ‘unseen’ colour differences. We need to be careful of telling people what they can or can not see 🙂 Happy shooting

          1. It is all just a matter of color filter array dyes and pigments as well as processing. nothing to do with CCD/CMOS.

        2. Charts and graphs are not, in themselves, “science”. I realize you are quite sure that these limited data points you reference validate your belief that “color difference is a myth”, but they are not sufficient to accomplish that.
          In the 70’s everyone believed that the best sounding amplifiers would be the ones with the lowest THD numbers, and, led on by the marketing departments pushing for ever lower THD numbers, the engineers obliged in the pursuit of vanishingly low THD numbers. Because it was a parameter which could be measured, as opposed to “how natural does this sound” which can’t be measured, the battle for lower THD numbers, analogous to the quest for ever bigger usable ISO numbers, went on for a long time, and gave us some of the best measuring and worst sounding amplifiers ever made.
          It’s like that.
          They say seeing is believing, but there are always those who are better at believing than they are at seeing.

          CMOS sensors obviously have certain advantages; those who, by dint of experience, are willing to forgo those advantages in order to gain a different look, that they can see, perhaps should be allowed to do that without being hectored by those who pretend to know what someone else can see, because they saw a graph on the Internet.

          1. Science of photoelectric effect tells there is no difference.
            Eric Fossum tells there is no difference (why not google who he is).
            Empirical test repeatable by anyone (which is science) tells there is no difference.
            DxOMark measurements of colour accuracy tells there is no CCD advantage.

            What can and is different is pigments (or dyes) of the colour filter arrays, especally with cameras from vastly different years.
            What can and is different is processing.

            I find it to be absurd that people tell it’s CCD (or CMOS) which makes the difference in their view even though the CFA and processing are unknown in subjective comparisons. Ignoring CFA and processing is just silly.

            “Different look” of CCD means mainly lower sigal to noise ratio. Just go and see how CCDs fare by DxOMark measurements. To simulate CCD on CMOS, just add some gaussian (or poisson) noise.

      2. Also, about CMOS advantages: far lower noise nowdays, higher full well capacity, higher quantum efficiency – all these mean higher image quality (and the first two mean larger dynamic range).

        Additionally CMOS sensors use much less power than CCD (+the required add on electronics not needed by CMOS). For live view purposes this is especially big thing.

        Also CMOS has natural anti-blooming protection – CCD and very bright lights tends to cause problems.

        Much much more money and resources has been put on CMOS developement over the last decade than on CCD developement and it shows. CCD is more or less dead for protography. Outside of this niche there are still uses, but not much because of performance.

          1. For photogaphic cameras CMOS is objectively speaking a better option. CCD developement has been almost non-existing more than a decade, while CMOS has gotten zillions of dublons in R&D money. 10 or 15 years ago you statement would have been relatively true, but now it’s not.

            If we were to have two cameras on the market, with no other difference (including cost of manufacture) than one having CCD based imaging subsystem and the other a CMOS based one, there would be these differences and many more:

            CCD would run out of battery first
            CMOS would have more functional live view and contrast detect AF
            CCD would be lacking in dynamic range and have poorer low light and shadow performance
            CMOS output could emulate CCD output by adding some gaussian noise to the image (no one would notice difference in double blind test)
            The colors would be identical apart from deep shadows and low light where CMOS would win
            CCDs tend to suffer from striking (bright light burning whole column in an image)
            Due to budget CMOS would have higher resolution

            There are lots of folks who prefer CCD, I am sure. However their preference is not based on anything objective, but either in subjective emotion, or thinking that the difference in some specific systems regarding some specific metric was because of CMOS/CCD when it really was because of something else (for example color is not because of CCD/CMOS, but because of processing and the materials used in the color filter array sitting on top of the sensor).

      3. You know that sensors, be they CCD or CMOS, only ‘see’ in black and white and that’s the RGB array, plus the software of the camera, that allow for colour rendition, don’t you?

          1. Actually not. See my reply to you earlier 🙂

            There is zero hard evidence which would say CCD has better (ie more accurate) colour. Zero.

            There are many subjective stories which ignore the real culprits of colour difference – differences in the CFA materials and image processing.

            On the other hand the science of photoelectric effect and scientific tests agree that there is no difference (apart from as per Eric Fossum, CMOS might have sligtly better blue response).

            Silicon is just silicon, wether the technology used to measure the number of electrons is CMOS or CCD. The capturing itself is identical.

  10. Hasselblad is rumoured to announce a new prosumer camera next week. A full frame camera with M mount at $4k would be competitive and it wouldn’t compete with their own medium format offerings. I think that’s the most likely scenario.

    It’s highly unlikely that it will come from Leica and if so, it will be priced considerably above the M240. I think it’s also very unlikely that it will come from Zeiss or Ricoh. Maybe from Cosina/Voigtländer, but I find that still unlikely. And Peter already said he wasn’t talking about the Konost. So maybe a new Hasselblad after all?

  11. Ok Peter….here is my final guess for a bit of fun.

    XPan digital using pano sensor at 50MP with crop function to 35mm (and maybe 1:1). So MP on these crops reduced accordingly. Would perhaps need native lens for pano but in crop use M Mount lenses.

    RF ….not sure. As you know the original XPan was a rangefinder so maybe but the way EVFs have come on wouldn’t surprise me if this only.

      1. Now THIS is a very good guess. CCD is used in some large sensor/medium format digital, where live view/video is not that important, but quality at base ISO. Here the CCD can perform its best. And that crop mode for M lenses + native for X-pan is a very good guess… I think it is doable. Even, digital rangefinders (for example one that continously checks the RF patch with an internal camera that allows for user-defined micro-adjustment, as the patch is not mechanical anymore) is also posible. Let’s see…

        1. Problem with that hypothesis is that CMOS beats CCD at low ISO as well. The reason why medium format often uses CCD is that it’s available – there aren’t much market for such speaciality chips, thus only recently a medium format sized CMOS appeared.

          Anyhow, if a sensor doesn’t do well at high ISO, it won’t do well at lower ISOs either. The key is SNR (signal to noise ratio). Every time you double the amount of light you capture you improve SNR by factor of 1.414 (ie sqrt2). When you shoot at lower ISOs you capture more light – this is why they are less noisy than high ISOs. Now, if sensor A is twice as noisy as sensor B at high ISO, it’s going to be twice as noisy at lower ISOs as well. (There is a small caveat regarding ADC, deep shadows and dynamic range, but thats for cameras with inferior ADC, like many (especially older) Canons – nothing to do with CCD/CMOS).

          CCD technology has just about only one advantage nowdays from performance point of view – TDI (time delay integration). For conventional photography this means that binning on CCDs can reduce the total read noise (ie improve deep shadow (and thus high ISO) performance at cost of reduced resolution). For scientific/industrial uses it has more practical advantages – for us not so much as CCDs drag behind in read noise department.

          Really hard to see why CCD would be used in any camera. Maybe someone has managed to get chips at low price. Or maybe there will be snapshot shutter instead of mechanical one, though that loses half the light sensitivity as well as FWC – no full frame CMOS on the market can do that.

  12. I hope they use the Dalsa 32MPixel CCD. This is in the same family as the CCD used by Phase 1 in their new Monochrome camera. Dalsa makes the 32MPixel CCD in color and monochrome. It has better dynamic range and better resolution than the KAF-18500 used in the M9 and M Monochrom. These CCD’s need to be in a Leica!

Leave a Reply to Chris Cancel reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s